app.py

import gradio as gr
import spaces
import os
os.environ["HF_HUB_ENABLE_HF_TRANSFER"] = "1"

import sys
sys.path.insert(0, './diffusers/src')

import torch
import torch.nn as nn

#Hack for ZeroGPU
torch.jit.script = lambda f: f
####

from huggingface_hub import snapshot_download
from diffusers import DPMSolverMultistepScheduler
from diffusers.models import ControlNetModel

from transformers import CLIPVisionModelWithProjection

from pipeline import OmniZeroPipeline
from insightface.app import FaceAnalysis
from controlnet_aux import ZoeDetector
from utils import draw_kps, load_and_resize_image, align_images

import cv2
import numpy as np

base_model="stablediffusionapi/clarity-xl"

snapshot_download("okaris/antelopev2", local_dir="./models/antelopev2")
face_analysis = FaceAnalysis(name='antelopev2', root='./', providers=['CPUExecutionProvider'])
face_analysis.prepare(ctx_id=0, det_size=(640, 640))

dtype = torch.float16

ip_adapter_plus_image_encoder = CLIPVisionModelWithProjection.from_pretrained(
    "h94/IP-Adapter", 
    subfolder="models/image_encoder",
    torch_dtype=dtype,
).to("cuda")

zoedepthnet_path = "okaris/zoe-depth-controlnet-xl"
zoedepthnet = ControlNetModel.from_pretrained(zoedepthnet_path,torch_dtype=dtype).to("cuda")

identitiynet_path = "okaris/face-controlnet-xl"
identitynet = ControlNetModel.from_pretrained(identitiynet_path, torch_dtype=dtype).to("cuda")

zoe_depth_detector = ZoeDetector.from_pretrained("lllyasviel/Annotators").to("cuda")

pipeline = OmniZeroPipeline.from_pretrained(
    base_model,
    controlnet=[identitynet, zoedepthnet],
    torch_dtype=dtype,
    image_encoder=ip_adapter_plus_image_encoder,
).to("cuda")

config = pipeline.scheduler.config
config["timestep_spacing"] = "trailing"
pipeline.scheduler = DPMSolverMultistepScheduler.from_config(config, use_karras_sigmas=True, algorithm_type="sde-dpmsolver++", final_sigmas_type="zero")
pipeline.load_ip_adapter(["okaris/ip-adapter-instantid", "h94/IP-Adapter", "h94/IP-Adapter"], subfolder=[None, "sdxl_models", "sdxl_models"], weight_name=["ip-adapter-instantid.bin", "ip-adapter-plus_sdxl_vit-h.safetensors", "ip-adapter-plus_sdxl_vit-h.safetensors"])

def get_largest_face_embedding_and_kps(image, target_image=None):
    face_info = face_analysis.get(cv2.cvtColor(np.array(image), cv2.COLOR_RGB2BGR))
    if len(face_info) == 0:
        return None, None
    largest_face = sorted(face_info, key=lambda x: x['bbox'][2] * x['bbox'][3], reverse=True)[0]
    face_embedding = torch.tensor(largest_face['embedding']).to("cuda")
    if target_image is None:
        target_image = image
    zeros = np.zeros((target_image.size[1], target_image.size[0], 3), dtype=np.uint8)
    face_kps_image = draw_kps(zeros, largest_face['kps'])
    return face_embedding, face_kps_image

@spaces.GPU()
def generate(
    prompt="A person",
    composition_image="https://github.com/okaris/omni-zero/assets/1448702/2ca63443-c7f3-4ba6-95c1-2a341414865f",
    style_image="https://github.com/okaris/omni-zero/assets/1448702/64dc150b-f683-41b1-be23-b6a52c771584",
    identity_image="https://github.com/okaris/omni-zero/assets/1448702/ba193a3a-f90e-4461-848a-560454531c58",
    base_image="https://github.com/okaris/omni-zero/assets/1448702/2ca63443-c7f3-4ba6-95c1-2a341414865f",
    seed=42,
    negative_prompt="blurry, out of focus",
    guidance_scale=3.0,
    number_of_images=1,
    number_of_steps=10,
    base_image_strength=0.15,
    composition_image_strength=1.0,
    style_image_strength=1.0,
    identity_image_strength=1.0,
    depth_image=None,
    depth_image_strength=0.5,
    progress=gr.Progress(track_tqdm=True)
):
    resolution = 1280

    if base_image is not None:
        base_image = load_and_resize_image(base_image, resolution, resolution)
    else:
        if composition_image is not None:
            base_image = load_and_resize_image(composition_image, resolution, resolution)
        else:
            raise ValueError("You must provide a base image or a composition image")

    if depth_image is None:
        depth_image = zoe_depth_detector(base_image, detect_resolution=resolution, image_resolution=resolution)
    else:
        depth_image = load_and_resize_image(depth_image, resolution, resolution)

    base_image, depth_image = align_images(base_image, depth_image)

    if composition_image is not None:
        composition_image = load_and_resize_image(composition_image, resolution, resolution)
    else: 
        composition_image = base_image

    if style_image is not None:
        style_image = load_and_resize_image(style_image, resolution, resolution)
    else:
        raise ValueError("You must provide a style image")
    
    if identity_image is not None:
        identity_image = load_and_resize_image(identity_image, resolution, resolution)
    else:
        raise ValueError("You must provide an identity image")
    
    face_embedding_identity_image, target_kps = get_largest_face_embedding_and_kps(identity_image, base_image)
    if face_embedding_identity_image is None:
        raise ValueError("No face found in the identity image, the image might be cropped too tightly or the face is too small")
    
    face_embedding_base_image, face_kps_base_image = get_largest_face_embedding_and_kps(base_image)
    if face_embedding_base_image is not None:
        target_kps = face_kps_base_image

    pipeline.set_ip_adapter_scale([identity_image_strength,
        {
            "down": { "block_2": [0.0, 0.0] },
            "up": { "block_0": [0.0, style_image_strength, 0.0] }
        },
        {
            "down": { "block_2": [0.0, composition_image_strength] },
            "up": { "block_0": [0.0, 0.0, 0.0] }
        }
    ])

    generator = torch.Generator(device="cpu").manual_seed(seed)

    images = pipeline(
        prompt=prompt,
        negative_prompt=negative_prompt,
        guidance_scale=guidance_scale,
        ip_adapter_image=[face_embedding_identity_image, style_image, composition_image],
        image=base_image,
        control_image=[target_kps, depth_image],
        controlnet_conditioning_scale=[identity_image_strength, depth_image_strength],
        identity_control_indices=[(0,0)],
        num_inference_steps=number_of_steps, 
        num_images_per_prompt=number_of_images,
        strength=(1-base_image_strength),
        generator=generator,
        seed=seed,
    ).images

    return images

#Move the components in the example fields outside so they are available when gr.Examples is instantiated


with gr.Blocks() as demo:
    gr.Markdown("<h1 style='text-align: center'>Omni Zero</h1>")
    gr.Markdown("<h4 style='text-align: center'>A diffusion pipeline for zero-shot stylized portrait creation [<a href='https://github.com/okaris/omni-zero' target='_blank'>GitHub</a>], [<a href='https://styleof.com/s/remix-yourself' target='_blank'>StyleOf Remix Yourself</a>]</h4>")
    with gr.Row():
        with gr.Column():
            with gr.Row():
                prompt = gr.Textbox(label="Prompt", value="A person")
            with gr.Row():
                negative_prompt = gr.Textbox(label="Negative Prompt", value="blurry, out of focus")
            with gr.Row():
                with gr.Column(min_width=140):
                    with gr.Row():
                        composition_image = gr.Image(label="Composition")
                    with gr.Row():
                        composition_image_strength = gr.Slider(label="Strength",step=0.01, minimum=0.0, maximum=1.0, value=1.0)
            #with gr.Row():
                with gr.Column(min_width=140):
                    with gr.Row():
                        style_image = gr.Image(label="Style Image")
                    with gr.Row():
                        style_image_strength = gr.Slider(label="Strength",step=0.01, minimum=0.0, maximum=1.0, value=1.0)
                with gr.Column(min_width=140):
                    with gr.Row():
                        identity_image = gr.Image(label="Identity Image")
                    with gr.Row():
                        identity_image_strength = gr.Slider(label="Strenght",step=0.01, minimum=0.0, maximum=1.0, value=1.0)
            with gr.Accordion("Advanced options", open=False):
                with gr.Row():
                    with gr.Column(min_width=140):
                            with gr.Row():
                                base_image = gr.Image(label="Base Image")
                            with gr.Row():
                                base_image_strength = gr.Slider(label="Strength",step=0.01, minimum=0.0, maximum=1.0, value=0.15, min_width=120)
                #    with gr.Column(min_width=140):
                #     with gr.Row():
                #         depth_image = gr.Image(label="depth_image", value=None)
                #     with gr.Row():
                #         depth_image_strength = gr.Slider(label="depth_image_strength",step=0.01, minimum=0.0, maximum=1.0, value=0.5)
                            
                with gr.Row():
                    seed = gr.Slider(label="Seed",step=1, minimum=0, maximum=10000000, value=42)
                    number_of_images = gr.Slider(label="Number of Outputs",step=1, minimum=1, maximum=4, value=1)
                with gr.Row():
                    guidance_scale = gr.Slider(label="Guidance Scale",step=0.1, minimum=0.0, maximum=14.0, value=3.0)
                    number_of_steps = gr.Slider(label="Number of Steps",step=1, minimum=1, maximum=50, value=28)
            
        with gr.Column():
            with gr.Row():
                out = gr.Gallery(label="Output(s)",format="png")
            with gr.Row():
                # clear = gr.Button("Clear")
                submit = gr.Button("Generate")
        
                submit.click(generate, inputs=[
                    prompt,
                    composition_image,
                    style_image,
                    identity_image,
                    base_image,
                    seed,
                    negative_prompt,
                    guidance_scale,
                    number_of_images,
                    number_of_steps,
                    base_image_strength,
                    composition_image_strength,
                    style_image_strength,
                    identity_image_strength,
                    ],
                    outputs=[out]
                )
        # clear.click(lambda: None, None, chatbot, queue=False)
if __name__ == "__main__":
    demo.launch()